We have measured the mechanical strength (sigma) of pure water ices V
and VI under steady state deformation conditions. Constant displacemen
t rate compressional tests were conducted in a gas apparatus at confin
ing pressures from 400 < P < 800 MPa, temperatures from 209 < T < 270
K, and strain rates 7 x 10(-7) < epsilon < 7 x 10(-4) s(-1). Most of t
he results fit to an empirical flow law of the form epsilon = A sigma(
n) exp (-(E + PV*)/RT), where the four material constants A, n, E*, a
nd V are (for epsilon in inverse seconds and P and cr in megapascals)
10(23.0), 6.0, 136 kJ/mol, and 29 cm(3)/mol, respectively, for ice V;
and 10(6.7), 4.5, 66 kJ/mol, and 11 cm(3)/mol, respectively, for ice
VI. Ice VI may weaken to a mechanism of higher E at T > 250 K. Ices V
and VI are thus theologically distinct but by coincidence have approx
imately the same strength under the conditions chosen for these experi
ments. To avoid misidentification, these tests are therefore accompani
ed by careful observations of the occurrences and characteristics of p
hase changes. One sample each of ice V and VI was quenched at pressure
to metastably retain the high-pressure phase and the acquired deforma
tion microstructures; X ray diffraction analysis of these samples conf
irmed the phase identification. Surface replicas of the deformed and q
uenched samples suggest that ice V probably deforms largely by disloca
tion creep, while ice VI deforms by a more complicated process involvi
ng substantial grain size reduction through recrystallization.